Hi!
We have a small production
ceph cluster, based on firefly release.
Now, client and cluster network shared the same IP range and VLAN. Also the same network
used for OpenNebula instance, that we use to manage our cloud. This network segment
was created some time ago and grows up with installation, but it have JumboFrames disabled.
Old network have 775 VLAN ID and 192.168.75.0/24 IP range.
We want to separate storage IO traffic into newly created VLAN/IP with JumboFrames enabled,
but ceph server ~90 VMs and we cannot stop the whole cluster. New network have 776 VLAN ID
and 192.168.76.0/24 IP range. It available on all machines in cluster as a failover bridge
(in active-backup mode, cause we also provide redundancy for our ceph nodes).
The question is: is it possible to migrate the whole cluster to a new nework segment
without downtime? What method you can recommend and can anyone provide step-by-step
instructions of such migration?
PS. Both network are separated at L2 level, so there is no routing between them, though we can
make them routable and reachable temporarily for migration. Also I think, that ceph highly relied
on node names, so I think there will be a some tasks to reconfigure DNS zones iteratively.
Thanks!
Megov Igor
yuterra.ru, CIO
megov@xxxxxxxxxx
Now, client and cluster network shared the same IP range and VLAN. Also the same network
used for OpenNebula instance, that we use to manage our cloud. This network segment
was created some time ago and grows up with installation, but it have JumboFrames disabled.
Old network have 775 VLAN ID and 192.168.75.0/24 IP range.
We want to separate storage IO traffic into newly created VLAN/IP with JumboFrames enabled,
but ceph server ~90 VMs and we cannot stop the whole cluster. New network have 776 VLAN ID
and 192.168.76.0/24 IP range. It available on all machines in cluster as a failover bridge
(in active-backup mode, cause we also provide redundancy for our ceph nodes).
The question is: is it possible to migrate the whole cluster to a new nework segment
without downtime? What method you can recommend and can anyone provide step-by-step
instructions of such migration?
PS. Both network are separated at L2 level, so there is no routing between them, though we can
make them routable and reachable temporarily for migration. Also I think, that ceph highly relied
on node names, so I think there will be a some tasks to reconfigure DNS zones iteratively.
Thanks!
Megov Igor
yuterra.ru, CIO
megov@xxxxxxxxxx
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
